We have earlier demonstrated for the first time that endocannabinoids acting via CB1 receptors promote voluntary alcohol drinking in an age-dependent manner, using a mouse model of two bottle/free choice paradigm(PNAS 100:1393, 2003). The brain-penetrant CB1 inverse agonist rimonabant used in this study was subsequently introduced as a treatment of obesity, but had to be wihdrawn from the pharmaceutical market in 2008, due to neuropsychiatric side effects, including anxiety, depression and suicidal ideation. An alternative approach, championed by my laboratory, was the introduction of peripherally restricted CB1 inverse agonists that retained the metabolic efficacy of rimonabant but were devoid of its neurobehavioral effects in rodent models of the metabolic syndrome. Paradoxically, such compounds were also found to reduce food intake, a centrally mediated effect. This paradox was resolved by the demonstration that peripheral CB1 blockade in diet-induced obese mice rapidly reversed their leptin resistance by reversing their hyperleptinemia, via inhibiting leptin production in adipose tissue and increasing leptin clearance in the kidney. This made us to wonder whether the high alcohol preference of C57Bl6 mice, another central function promoted by CB1 activation, may also be affected indirectly through blockade of CB1 in the periphery. A possible mechanism involves ghrelin, a gastric peptide that promotes appetite via ghrelin receptors in the brain. The preliminary findings reported last year have been extended and the findings clearly indicate that rimonabant and the non brain-penetrant CB1 inverse agonists, JD5037, as well as several of our own novel peripheral CB1 antagonists including MRI-1569 and MRI-1891, were equieffective in markedly reducing total alcohol intake as well as ethanol preference in C57BL6 mice, using a two bottle, free choice paradigm. Furthermore, CB1 blockade significantly reduced plasma levels of total as well as acetylated ghrelin. In additional experiments conducted this year, we found that both alcohol preference and absolute intake are lower in ghrelin knockout and ghrelin receptor1 knockout mice, with no additional reduction caused by peripheral CB1 receptor blockade. The effects on plasma ghrelin and the finding in the knockout strains are compatible with ghrelin involvement in the effects of peripheral CB1 blockade on alcohol drinking. We also found that subdiaphragmatic vagal denervation of the stomach abolished the efficacy of peripheral CB1 blockade to reduce drinking. This suggests that presynaptic CB1 receptors on either efferent or afferent vagal terminals are the targets of these antagonists. Additional experiments in progress are to distinguish between these latter two possibilities.